Social Media Is Media Now
My student said social media wasn’t really social anymore. The research suggests the social didn’t disappear—it just moved out of the feed.
I launched this newsletter last week, which is a strange thing to do when you haven’t written for a general audience in years.
In a previous life, before academia took over most of my sentences, I worked as a journalist and wrote feature stories. But academic writing is a different kind of training altogether: careful, qualified, densely sourced, and usually addressed to a small group of people who already speak the language. Writing for the public again—more directly, more personally, and without hiding behind several pages of literature review—is something I’m still relearning. So bear with me while I find that voice again.
I could offer several explanations for why I stopped writing: the 4–4 teaching load, service and committee work, the tenure clock, and becoming a mother. They would all be true, but none of them would quite capture what the quiet felt like from inside.
Writing simply stopped feeling like talking to someone and started feeling like a task: something with a deadline, a reviewer, and a version of myself that had to show up correctly.
Then I spent several months reading research about why social media users seem to be posting less, and I kept recognizing something. A blank status-update box and a blank Word document are obviously not the same problem, and I’m not going to disguise my tenure clock as a structural critique.
But the reasons people give for going quiet kept sounding familiar: the bar for saying anything keeps rising, someone is always watching, and being wrong in public has become increasingly expensive.
Anyway. This is the piece I’m coming back with.
Last semester, a student in my online synchronous Computer-Mediated Communication class made an offhand comment that’s followed me around ever since.
We were discussing online self-presentation, i.e., how people decide what to share, what to hide, and which version of themselves appears on a screen, and the student unmuted to say, almost casually:
“Social media isn’t really social.”
Someone else agreed in the chat, and I remember saying something professor-like and kept the class moving. But the comment stuck.
At first I heard it as a description of the familiar frustrations of being online: too many fakes, too much political shouting, too many influencers insisting that a particular water bottle will reorganize your entire life. After several months of reading, I think he was pointing at something structural.
Social media hasn’t just become more annoying, or even overwhelming. What may have begun as a set of platforms organized primarily around interpersonal connection has increasingly become a system of broadcasting, entertainment, advertising, and passive consumption.
And yet we still call it social media, even though much of what now fills our feeds is barely social at all.
The Algorithm “Ate” Your Friends
I’m old enough to remember things back when logging onto social media meant seeing what people you knew were doing. A blurry photo of what someone had eaten, a complaint about work, an unedited vacation picture, or a status update that probably should’ve remained inside the person’s head.
The content wasn’t always interesting, but it was definitely personal.
Today, people are posting less, and I see this shift in my own classes. At the start of every semester, I ask students about their social media habits, i.e., what platforms they use, how often they post, and what they actually do once the app is open. The answers are remarkably consistent. Most use social media every day, for hours, but relatively few describe themselves as active posters. They scroll, they watch, they send memes to the same four people. Plenty of them simply don’t post at all.
I also see it in my own network, and Instagram stories are an unusually honest place to see it, because they hand you a receipt. Post one and you get a list: dozens of names, scrolled past in the ten seconds between other things. Colleagues, cousins, people I went to graduate school with, people I haven’t spoken to in years. They’re there. They’re watching. Almost none of them post anything themselves. My feed is quiet yet my viewer list isn’t, and those two facts describe the same people.
It’s a strange kind of intimacy—being seen by everyone you know while none of them are visible to you. It’s also, I’d argue, the whole phenomenon in miniature. There’s an audience. There’s just no exchange.
My classroom surveys aren’t representative research, and my viewer list is a sample of one. Still, this pattern closely resembles what larger studies are finding. A recent Ofcom report found that the percentage of adults in the United Kingdom who actively post on social media fell from 61 percent to 49 percent in a single year (McMahon, 2026). In the United States, a widely circulated survey reported more than half of respondents post less frequently than they did five years ago. Many respondents also noted that maintaining an online presence now feels “like work” (Rubenking, 2026).
The word work here is what stays with me. People are using the language of labor to describe something they supposedly do voluntarily [if not mindlessly], at times, and supposedly for fun. Which is exactly what I was describing three paragraphs into this essay—posting stopped feeling like talking and started feeling like a manuscript with a deadline, a reviewer, and a correct version of yourself to perform.
When I ask students why they post less, one answer comes up repeatedly: the algorithm. They’d describe feeds that are no longer feeling organized around the folks they chose to follow. Instead, they’re seeing recommended videos, sponsored posts, viral content, and creators they’ve never even looked up or heard of. While the language may often be blunt, e.g., “the algo doesn’t show me my friends anymore”, the diagnosis is quite accurate.
Part of the explanation really is the algorithm. Social media feeds aren’t primarily organized around interpersonal engagement anymore. They’re optimized for watch time, retention, and repeat consumption. Algorithms no longer prioritize content from people we actually know and instead prioritize whatever keeps you watching the longest, which essentially are polished videos from professional creators and, increasingly, artificial intelligence (AI)-generated or AI-assisted content, deemed more valuable to platforms than casual posts of ordinary users (Mir, 2026).
Our feeds now resemble algorithmic television channels with a like button rather than digital town squares (Goodwin, 2026; Laurenson, 2026; Mir, 2026).
Welcome to Parasocial Media
As creators replaced friends at the center of our feed, the relationships these platforms produce have also shifted too. danah boyd (2026) and Jess Maddox (2025) have both argued that social media may no longer be an accurate term when most of our time is spent watching strangers rather than interacting with peers. What we have instead is parasocial media.
Parasocial relationships are one-sided relationships that are more of a mediated bond in which someone develops a sense of familiarity or intimacy with a public figure who cannot know or reciprocate at the same level (Maddox, 2025; Ouvrein, 2026). This term came about through the work of Donald Horton and R. Richard Wohl (1956) to explain why audiences felt attached to a television celebrity, radio personality, or movie star.
What’s changed about the dynamics today is perhaps the delivery. Influencers look directly into a phone camera and share intimate, behind-the-scenes details of their lives (Maddox, 2025; Ouvrein, 2026).
“You need to hear this.”
“Come with me.”
“Let’s get ready together.”
Filming from bedrooms, kitchens, cars, and hospital waiting rooms; sharing breakups, diagnoses, pregnancies, unfiltered hot takes. The content feels personal because it’s been designed to feel personal.
Here’s the part that interests me most, and it’s where my own corner of the field has something to say.
When we assess people we actually know, we lean—often without noticing—on information that would be hard for them to fake. Communication researchers call this a cue’s warranting value, and the underlying principle is that the less control a person has over information about themselves, the more we treat that information as true (DeAndrea, 2014). A photo someone else tagged you in, a mutual friend’s offhand comment, the fact that you showed up somewhere and other people saw you: these carry weight precisely because you didn’t manufacture them. Information you generate about yourself carries almost none.
Influencer intimacy is made almost entirely of the second kind. The messy kitchen, the no-makeup morning, the phone propped against a mug at an artfully casual angle —every one of those cues is self-generated, which is to say, low-warrant by definition. But the form of the address mimics high-warrant intimacy. Being close to someone’s unglamorous life is normally something you earn through proximity and time. The direct-to-camera format simulates the output of that process without any of the process, and our warranting instincts—calibrated over a lifetime of embodied social life—don’t reliably distinguish the two.
That’s the trick. Not that influencers lie or are inauthentic; it’s just that the platform delivers self-presentation in some kind of wrapper that our brains read as evidence.
Maddox (2025) pushes the argument further by suggesting that we’re not only becoming parasocial toward celebrities, we’re also becoming parasocial toward each other. At the far end, this means anyone with a moderately popular TikTok becoming an object of familiarity for thousands of strangers who joke with them, critique their appearance, and demand explanations as if they’re old friends. It doesn’t take thousands for this effect to kick in; however, all it takes is a[n] [Instagram] story or reel, and a viewer list.
Going back to mine: forty-seven people receive a piece of my self-presentation, returning nothing I could use to assess them, and knowing something about my week that I know nothing about theirs. And it’s the same asymmetry, running at a scale so ordinary that we’ve stopped noticing it’s an asymmetry at all.
The line between public figure and ordinary person is collapsing, and it isn’t collapsing because ordinary people got famous. It’s collapsing because the ordinary version of being watched now has the same shape as the famous one.
Almost anyone with a visible online presence can now become a character in someone else’s media environment.
Parasocial is a Diagnosis, Not an Insult
Curiously, parasocial has become somewhat of a slur. It’s a way of telling someone their attachment is embarrassing, delusional, or fake.
An influencer doesn’t even know you exist.
Technically, while this may be true, it misses what makes parasocial relationships feel incredibly powerful—even providing real comfort—for some.
For an international student a thousand miles from home at 2 a.m., a familiar voice narrating a familiar routine is a sensible way to feel less alone. For someone who has never seen a person like themselves treated as ordinary and worth listening to, watching someone like themselves be ordinary in public can be genuinely clarifying, even gratifying.
What’s new about parasociality isn’t the attachment; it’s that the attachment itself has been industrialized. While Horton and Wohl (1956) were describing an accident of broadcasting—a side effect of a camera pointed at a person—we’re now describing a design goal. Parasocial bond strength is an engagement metric and metrics get optimized. The implication then is that when we call a viewer parasocial as an insult, we’re blaming someone for responding as designed to a system built to produce exactly that very response. In effect, we’re letting the design itself off the hook.
The cost shows up in the research, carefully. Ouvrein (2026) found a positive association between stronger parasocial ties to influencers and perceived loneliness. One explanation is displacement, i.e., time and emotional energy spent tracking an influencer’s life leave less room for the reciprocal kind.
We all know correlation isn’t causation. Strong parasocial attachment may contribute to loneliness—people who are already lonely may also be more likely to seek companionship from influencers. Nothing stops both from being true at the same time.
Notably, Ouvrein found real friendships mediate the relationship, which suggests the question isn’t whether parasocial ties exist but what else is in your life alongside them.
And herein lies the tension.
Parasocial media gives you many of the sensations of connection but [conveniently] without the responsibilities of a relationship. While parasocial ties may supplement our existing relationships, they become more concerning when they begin to substitute for the inconvenient work of showing up for people who can also ask something of us.
When Socializing Went Behind The Door
Despite data showing that people are posting less publicly, it doesn’t mean they’ve stopped communicating online altogether. People didn’t just stop socializing.
What’s instead happening is that the socializing went underground. Or, in this case, it’s shifted to a place with a door, so to speak.
As public feeds increasingly feel performative, exhausting, or even toxic, social media users are migrating to what Pranav Malhotra (2026) calls bounded social media places (BSMPs). These refer to restricted, lower-visibility spaces that give users more control over who enters and who can see what they share (Malhotra, 2026).
Today, our mediated socializing is increasingly showing up more in “the group chat” that has somehow remained active for nine years despite nobody remembering who had created it. Think of our WhatsApp groups, private Instagram stories, direct-message threads, invitation-only Discord servers, close-friends lists, and anything in between.
In these environments, people feel like they’re able to speak to the audience they actually intended to reach, avoiding what scholars like Alice Marwick and danah boyd (2011) have called context collapse, i.e., the flattening of friends, relatives, coworkers, professors, exes, and acquaintances into a single undifferentiated audience. Rather than calculating how a post might read to everyone they've ever met, social media users retreat to groups whose expectations they understand.
Additionally, BSMPs are seen as spaces that are less exposed to recommendation algorithms, public search, unwanted spectators, and the pressure to make every post perform for an invisible crowd.
And so, it’s not that social media is becoming less social; it’s just that it’s becoming less public, with people hiding in these more private, bounded spaces so they feel they can be themselves without having to “perform” for an audience of strangers (Fisher, 2026; Malholtra, 2026; McMahon, 2026).
If the feed has become a stage, friendship has moved backstage.
For many social media users, especially younger ones, this migration to BSMPs to avoid context collapse is a rational, almost instinctive response. Private group chats, close-friends lists, and invite-only communities allow folks to communicate without constantly wondering who might see, misunderstand, or redistribute what they post.
In this sense, bounded spaces effectively restore some of the context that public platforms have gradually stripped away over the years. It’s remarkable how it took a little over a decade of context collapse to make people want their walls back.
The Private Room isn’t a Complete Solution
While it may be tempting to end with a cheerful recommendation of abandoning the public feed, joining a group chat, and rediscovering “authentic” connection, the fact of the matter is that moving sociality into private spaces, too, comes with certain tradeoffs. A group chat may support friendship, but it doesn’t necessarily support public life.
Engaging communities based on exclusion may be comforting to some, but can also make things harder for others to enter, particularly for people who don’t already have strong networks elsewhere. Ironically, the person who most needs a room to be let into is the person least likely to be handed the invite link.
The same goes for the more appealing version of retreating from the internet-mediated realm altogether. As some of us may have observed from our own social network feeds, some users are responding to platform fatigue by logging off and rebuilding physical social infrastructure, i.e., community gardens, book clubs, neighborhood groups, face-to-face organizing, essentially places where relationships are built through repeated presence and mutual obligation (Fisher, 2026; Love, 2026). I think this is genuinely good, but I want to be careful about what it costs.
Making real life more attractive isn’t a matter of personal discipline. Community requires time, transportation, affordable places to gather, childcare, accessible public space, and institutions people can actually trust.
Real life has a budget.
When public infrastructure degrades, people with resources leave and build smaller private alternatives, and then everyone else stays in the degraded public. The public option declines even further and private escape becomes somewhat of a lifestyle brand. And then—this is the part that bothers me—the ability to “escape” gets mistaken for personal virtue. Deleting the app becomes a character trait rather than a resource.
We have watched this happen with schools, with parks, with public transit. There’s no obvious reason the feed is exempt.
There’s a research problem here too, and it’s mine as much as anyone else’s. While public social media is visible, private messages, closed groups, and invitation-only servers are not. If the meaningful interaction has moved somewhere we can’t see, we’ll keep producing increasingly precise accounts of the part of digital life that’s easy to access, and increasingly confident theories about a public feed that fewer and fewer people are actually using for anything social. We’ll be excellent scholars of the empty room.
The Feed is Media Now
This brings me back to where we started. My student may have been right—at least about our public-facing feed. What we still call social media is increasingly functioning as a broadcasting and entertainment system where people discover recipes, follow breaking news, encounter new ideas, watch funny videos, and receive surprisingly strong opinions about skin care from a 22-year-old with excellent lighting.
I’m not suggesting that any of that is useless. It simply means that perhaps we should stop expecting the feed to carry our friendships for us (Dickson, 2026). Especially when the reality has always been that the feed is designed to keep us watching, giving us people to recognize, when actual real friendship requires us to actively participate and involves people who actually recognize us back.
Perhaps one of the most important digital literacy skills nowadays isn’t learning how to quit social media entirely—it’s learning to distinguish between media consumption and genuine social connection, all the while noticing when a system has been designed to blur it. Even though both may have their respective values, they’re not the same activity.
And so, it’s not that the social part of social media has vanished completely; it’s just that it’s been pushed out of our feed and has instead moved into private messages, group chats, closed communities, and offline spaces. It's basically migrated to places where people can still see one another without first having to satisfy an algorithm.
Going around the feed rather than through it might help us hold on to our relationships. I’m just not sure what happens to everything we leave behind along the way.
Where are you finding actual community right now—online or offline? And, more importantly, who isn’t in the room when you find it?
Read Alongside
boyd, d. (2026). Social media is now parasocial media. Social Media + Society, 1-5. https://doi.org/10.1177/20563051261437487.
DeAndrea, D. C. (2014). Advancing warranting theory. Communication Theory, 24(2), 186–204. https://doi.org/10.1111/comt.12033
Fisher, L. (2026, April 21). Social media doesn’t feel social anymore – so where does online community go from here? MIDiA Research. https://www.midiaresearch.com/blog/social-media-doesnt-feel-social-anymore-so-where-does-online-community-go-from-here
Goodwin, T. (2026, February 17). How social media stopped being social. The Drum. https://www.thedrum.com/opinion/tom-goodwin-how-social-media-stopped-being-social
Horton, D., & Wohl, R. R. (1956). Mass communication and para-social interaction: Observations on intimacy at a distance. Psychiatry, 19(3), 215–229. https://doi.org/10.1080/00332747.1956.11023049
Laurenson, J. (2026, May 22). Anti-social: It's fads, not friends, which now dominate our feeds. BBC News. https://www.bbc.com/worklife/article/20260520-how-social-media-ceased-to-be-social
Love, S. (2026, July 5). The moral case for being less online. Vox. https://www.vox.com/future-perfect/494129/quit-social-media-political-influence-offline-luddite
Maddox, J. (2025, December 1). BTW: We are all parasocial now. BTW.
Malhotra, P. (2026). Toward conceptualizing bounded social media places. Political Communication, 43(2), 265–281. https://doi.org/10.1080/10584609.2025.2512342.
Marwick, A. E., & boyd, d. (2011). I tweet honestly, I tweet passionately: Twitter users, context collapse, and the imagined audience. New Media & Society, 13(1), 114–133. https://doi.org/10.1177/1461444810365313
McMahon, L. (2026, April 2). Fewer UK adults posting on social media, Ofcom finds. BBC News. https://www.bbc.com/news/articles/cgqk718l4neo
Mir, A. (2026, July 10). How social media became antisocial. City Journal. https://www.city-journal.org/article/social-media-posts-algorithms
Ouvrein, G. (2026). I’ll be online for you: Parasocial relationships with social media influencers, loneliness and the mediating role of real friendships. Current Psychology, 45, 198. https://doi.org/10.1007/s12144-025-08630-0.
Rubenking, N. J. (2026, July 11). The death of the status update: Why 55% of Americans stopped posting on social media. PCMag. https://www.pcmag.com/news/death-of-the-status-update-why-55-americans-stopped-posting-social-media
